Network News

X My Profile
View More Activity
The new Washington
Post Weather website
Jump to CWG's
Latest Full Forecast
Outside now? Radar, temps
and more: Weather Wall
Follow us on Twitter (@capitalweather) and become a fan on Facebook
Posted at 1:15 AM ET, 12/23/2010

Late night storm thoughts

By Jason Samenow

I'm trying something new tonight: I'll post quick thoughts as the different computer models stream in as well as other odds and ends (about every 15-30 minutes). Keep your finger near the refresh button...

1:15 a.m. update: European model a tad more out to sea and not as intense as last several runs. Would be a big hit for the coast, but just light snow for us. Little support now for a super snowstorm in D.C. Pretty huge for New England though. Right now, (considering everything) the forecast we've held onto since basically Sunday for a 50-60% of 1" or more, and 20% or so chance for more than 4" seems about right. And the possibility of a miss is still out there.

12:50 a.m. update: European model starting to come out. Hang tight.

12:20 a.m. update: The GFS ensemble mean (the average of the various versions of the GFS model runs with slightly different initial conditions) is a bit west of the primary operational run and produces a little more precipitation.

12:00 a.m. update The Canadian model tracks everything too far to the south and east and the D.C. metro region gets nada. I'm not a big fan of the Canadian model, but it definitely shows we're not near a consensus about what's going to happen.

11:30 p.m. update: Another takeaway from the models coming in so far tonight is that they are unified in at least giving the mid-Atlantic some snow and phasing the northern and southern stream energy - albeit fully a little late for a big snow here. But the details in the timing/phasing will still evolve so my confidence is a little higher we'll get some snow given the direction of the models. Plus, still awaiting the Euro model coming out at 1 a.m. Who's staying up???

11:15 p.m. update: Both the GFS and the NAM suggest a big storm will develop but not until it passes north and northeast of D.C. The dip in the jet stream in these models doesn't amplify enough for the storm to come up the coast in the mid-Atlantic - probably because they advance the southern stream too quickly and not as far south. However, both models would suggest some light accumulating snow would fall late on Christmas day into early Sunday morning.

11:00 p.m. update: GFS run is also speeding up the southern stream and bringing back the threat of snow on Christmas. But it doesn't look like it's going to develop the storm explosively until it passes north of D.C.

10:50 p.m. update: Another perspective - Henry Margusity at AccuWeather is predicting 6"+ for the metro region...

10:45 p.m. update: One thing about the NAM model is that it does look like it's primed to bomb (develop rapidly) the low after 84 hours - so it does suggest a big storm will form - just a little late for DC, but maybe not so further north. But at this time frame, it's very close. And this model - again - is not real skillful at this timeframe.

10:20 p.m. update: One meteorologist, Dave Tolleris, at has released some snowfall probabilities for the East Coast. You can see them here. He's pretty bullish about the storm threat, indicating a 50-70% chance of 4-8" for the metro region. Ours aren't that high (20-30% for 4" or more) but we'll be updating them tomorrow after we digest tonight's models and tomorrow morning's.

10:00 p.m. update: The latest NAM model is in - this is a short-range model that goes out just 84 hours. It actually brings back the idea of light snow on Christmas day as it advances the southern stream system more quickly. By the end of its run (84 hours, Sunday morning), it's trying to develop a storm off the coast of the Carolinas as the southern stream system interacts with a disturbance diving in from the northwest, but probably too far to the east to come up the coast and produce significant snow. I don't think this model, which is pretty poor with its forecasts beyond 48 to 60 hours, has a lot of support. Still, we'll need to see if its quicker timing is shown in other models coming out.

By Jason Samenow  | December 23, 2010; 1:15 AM ET
Categories:  Updates  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: California storm may bring South Christmas snow
Next: Forecast: Windy but plenty of sun

Comments is back up and his forecast is very adamant about the potential of a significant snow fall and the validity of the euro model to date with regards to the "pending" storm. is going too far out on a limb?

Posted by: ArlingtonWeather | December 22, 2010 10:13 PM | Report abuse

I like this rapid fire idea, at least for potential significant storms.

After all, this is a weather enthusiast's blog! :)

Posted by: JTF- | December 22, 2010 10:14 PM | Report abuse

I have to say, that wxrisk guy sounds a bit excitable.

Posted by: JTF- | December 22, 2010 10:21 PM | Report abuse


Were you reading my mind? I was posting about as you posted that comment... Tolleris is excitable, but a talented forecaster...he's very big on this storm potential. We're a little more guarded.

Posted by: Jason-CapitalWeatherGang | December 22, 2010 10:23 PM | Report abuse

Noooo, this means I'll be staying up till 'one more update' over and over... But thanks! And my fingers are crosed for snow :)

Posted by: RASinDC | December 22, 2010 10:30 PM | Report abuse

OK so at this point, I leave Ohio Sunday or Tuesday...

Posted by: SPS1 | December 22, 2010 10:30 PM | Report abuse

Jason, I've seen you note this twice, that is a "talented" forecaster. Your context is tantalizing. What do you mean by this? (If you addressed this before, just point me to the correct blog). thanks!

Posted by: weathergrrl | December 22, 2010 10:32 PM | Report abuse

I like this idea a lot--it means I don't have to wait up for one final "night report" before bed. Keep the updates coming.

Posted by: KBurchfiel | December 22, 2010 10:32 PM | Report abuse

"The December 25 - 27 Noreaster Winter Storm." :) wxrisk is not lacking flair

Posted by: Finn1917 | December 22, 2010 10:33 PM | Report abuse

jason - not trying to read minds. a little late in the evening for such activities. just trying to figure out the seemingly divergent forecasts of late between local media which seem to give the storm a low probability, cwg in its balanced approach, and being near certain of what will happen. at the end of the day, it appears that forecasting is based upon an initial view of a certain time period going forward based upon past experiences

Posted by: ArlingtonWeather | December 22, 2010 10:55 PM | Report abuse

Jason: reposted here, since you are on the clock tonight. Thanks very much -

What I find fascinating is the spread between the models as to where the low forms. Here's a pretty picture from 1993 that has a postion of a low somewhat in the middle. Q to the guys in charge, is there enough energy in CA, forward and cold air to make something like this happen again. I'm just entranced by the amount of energy out west.

Posted by: Gooddogs | December 22, 2010 10:57 PM | Report abuse

Oh gawd, I'm listening to the American wx radio show and hearing things about the 0z GFS I can't even see yet, the tension is killing me

Posted by: cleombrota | December 22, 2010 11:00 PM | Report abuse

Further confounding things is the apparent divergence within the local TV mets. Some are still in the conservative camp (V. Johnson, Ryan) and those who seem to be hinting at something bigger (Sader? From 9 and guy from channel 5).

Posted by: lingering_lead | December 22, 2010 11:01 PM | Report abuse

I'm curious to know how far out CWG considers models to have real value and worthy of discussion. I live in Minneapolis and forecasters are already talking about the possibility of a big storm here between Dec. 31 and Jan. 3, citing a GFS model (see also: and scroll down a bit). Do you think this is responsible reporting or sensationalizing this far out? Look forwarded to hearing your thoughts.

Posted by: eddiedog2 | December 22, 2010 11:01 PM | Report abuse

hello Jason. Do you know how much precip the
latest run of the GFS would give us?

Posted by: BH99 | December 22, 2010 11:06 PM | Report abuse

Joe Witte leans towards the out to sea option.

Posted by: lingering_lead | December 22, 2010 11:07 PM | Report abuse

Jason - I follow CWG religiously, and read the wxrisk post as well, and here's my question: I'm sure the GFS is a great asset, but isn't there a point where it confuses matters more vs. gives direction? Tolleris' point that it had the storm over Bermuda not that long ago, and the update now has it moving up a day makes it seem that is not grounded. A better way to say this is: if the Euro is wrong, you all have given me a great understanding as to what the factors might be, because it has been consistent - but if the GFS is right/wrong/sideways, I would have no idea, as it has seemed to say everything outside of rain of toads from here to Fiji in the last 48.

TL:DR - is there ever a point where you decide a model, however worthwhile historically, is hopeless with a specific event?

Posted by: awshux | December 22, 2010 11:20 PM | Report abuse


Looks like 0.1-0.25"

Posted by: Jason-CapitalWeatherGang | December 22, 2010 11:20 PM | Report abuse

Are we going to get blasted or not? End of story! I can't handle the stress anymore!

Posted by: greg2010 | December 22, 2010 11:21 PM | Report abuse

I am listening to the AWx radio thing that Wes is currently on and I keep hearing the other commentators say that they don't want to use the word "trend" b/c Wes doesn't like it. Can I get a classic CWG, user-friendly and insightful explanation of why Wes doesn't like the term "trend"?

Thanks guys!

Posted by: amandap2986 | December 22, 2010 11:25 PM | Report abuse

thx jason

Posted by: BH99 | December 22, 2010 11:33 PM | Report abuse

Re: 1130 update, will be here all night for all updates from now till 0630 tomorrow. Look forward to the updates Jason.

Posted by: Havoc737 | December 22, 2010 11:36 PM | Report abuse

i'd like to echo @awshux's question....

At what point do you start putting more faith in one model over another. since the EURO model has been pretty consistent over the last number of runs and the GFS is so scatter-brained does that, as a forecaster, give you more faith in the euro and less in the gfs?

is one of these models better at predicting east coast winter storms than the other?



Posted by: dotjeffdot | December 22, 2010 11:38 PM | Report abuse


I think Wes has probably gone to bed, but you could ask him in one of his upcoming posts.

Posted by: Jason-CapitalWeatherGang | December 22, 2010 11:39 PM | Report abuse

@dotjeffdot and @awshux

The Euro is a slightly more skillful model than the GFS beyond 72 hours from my understanding. We look at what models are most consistent, what models seem most physically realistic, what models seem to fit the pattern we're in, and what model first might simulate certain features that other models gradually all grasp onto.

Right now is seems like the Euro is the most consistent, leading model that other models are gradually getting closer to but are not all the way there. And there is a feeling that it's probably too extreme.

Posted by: Jason-CapitalWeatherGang | December 22, 2010 11:46 PM | Report abuse

Jason, I'll be here at 1 AM! Everyone at work asks me about this kind of stuff (since I'm a met by education, anyway) and assumes that I somehow still have time to look at every model and MOS product...have to get my info from somewhere, right?

Question - am I correct that Euro validated the most accurately (in the mid-Atlantic) last year, or is that just something I made up so my colleagues think that I know what I'm talking about?

Posted by: joshua-frank | December 22, 2010 11:51 PM | Report abuse

@Jason - thanks for the response. What I'm trying to get at is whether there is ever a point (take Snow-Ho-Ho out of it) where you just discount a model all together for a specific event because it has been too erratic on a run-by-run basis. Lastly, my alternate name would be "Blizzmas"

Posted by: awshux | December 23, 2010 12:03 AM | Report abuse

I do a ton of reading, sifting through models, listening to the TV mets and have lived here over 30 years (even took several climateology classes in college). I rarely make any comments, but what I will say having digested all available information over the past 5 days regarding this potential storm is the odds of a significant winter storm are going up when taking into account the evolution of all available data. Hard to say exact precipation amounts at this point, but all indications are the GFS is beginning to correlate itself to what the Euro has been saying now for nearly 2 days. If the 1am Euro run stays the course and tomorrow's GFS' continue to hedge westward then the models will be showing some modular consensus. Combine this with the Low off the New England coast beginning to shift eastward creating a nice atmospheric pocket for the 'Sunday to Monday' storm to enter and ride unhindered and unblocked up the eastern seaboard.

Posted by: TopperTed | December 23, 2010 12:03 AM | Report abuse


Yes-we will throw models out if they seem physically implausible and have no support...but that tends to happen more in the longer range because most of the models are decent inside 48 hours.

Posted by: Jason-CapitalWeatherGang | December 23, 2010 12:09 AM | Report abuse

Okay, before I turn in for the night, I want to confirm:
Are we still looking at a Sunday/Monday deal, or is the possibility (key word) of having some snow late on Christmas night still in the realm of possibility?

I too am the 'resident weather geek' for my office, and my completely unprofessional opinion is: We will either have a 'bust' of epic proportions, or we will get TOTALLY slammed. Based on the observations I've read today I don't see a lot of gray area here.

I've recommended to my parents when they go grocery shopping tomorrow that they prepare for 7 days (vice their normal 2-4 days). Also recommended they check their medications, etc. just in case.

My in-laws were planning to come up this way from NC (near Myrtle Beach) on Christmas Day, but have moved up their arrival here to MFY (my front yard) by one day. So the good thing is, they will probably see my kids open presents Christmas morning. A silver lining, in my book. :)

Whether my house will look like a Cat-5 hurricane hit it or not when they arrive, well that's another story for another time.

Goodnight all, and have a pleasant tomorrow. I'll be anxious to see updates!

Posted by: southbridgemom | December 23, 2010 12:19 AM | Report abuse

Correct me if im wrong here but with the Canadian being too far OTS it is not alone as the Navy model is as well, is that correct?

Any chance these 2 models are right over the Euro and GFS or is that impossible in that it has never happened?

This just wreaks Deja vu from last weekends storm. Hope you guys will update here about tonights Euro and its track and give your thoughts up to this in what track you think this storm may take with all models taken into consideration.

Im no met at all but just for the heck of it, im officially predicting the DC metro gets 0-3 inches from this storm. I just dont think it will be able to make that turn (close enough to us) the way some of these models are thinking.

I really want to hear all the CWG thoughts and opinions on this storm and if you think its at least in the bag at this point that we will see some snow. Im not asking for amounts, im just asking opinions on if we will see any?.... Maybe go out on a limb ;)

Posted by: KRUZ | December 23, 2010 12:21 AM | Report abuse


We're going to re-introduce the slight chance of some light snow Christmas night into the forecast, but the main event - should it materialize - would be Sunday (possibly into Monday).

Posted by: Jason-CapitalWeatherGang | December 23, 2010 12:26 AM | Report abuse

Im interested to see what the Euro will show at 1 am. Hopefully the model reaffirms the latest rumblings that DC could see 6 + inches :)

Posted by: Hailtotheredskns | December 23, 2010 12:29 AM | Report abuse

Heh, so we could end up with anything from a dusting to a foot? I like that range!

It seems like this winter has so far been a lot more difficult to forecast than last winter (which I don't recall as being particularly "easy" to forecast either). Props to CWG for guiding us through all the uncertainty.

Posted by: nlcaldwell | December 23, 2010 12:30 AM | Report abuse


I don't typically consider the Navy model or spend the time to look at it. It's just not very good and doesn't provide much value added over the others. Honestly, the most important models to look at are the Euro and its ensembles, the GFS and its ensembels, the NAM (short range), something called the SREF (short range), and sometimes the Canadian and/or the UKMET.

Will stay up for the Euro and let you know my thinking at that point.

Posted by: Jason-CapitalWeatherGang | December 23, 2010 12:30 AM | Report abuse

@Jason-- Thanks so much! Will advise the family there is still a chance of precip Christmas night. Trying to keep safety and sanity at the forefront of holiday activities/decisions.

THIS is why CWG is my 'one stop' source for wx info for DC.

Now where is my pillow?!!?

Posted by: southbridgemom | December 23, 2010 12:33 AM | Report abuse

Jason, amazing work here. I think you're going above and beyond the normal call of duty tonight, and I and countless other readers appreciate your time and expertise!

Posted by: natsncats | December 23, 2010 12:35 AM | Report abuse

Over time the Euro seems most consistent...and it's the model which seems the most bullish for us on this storm.

Posted by: Bombo47jea | December 23, 2010 12:48 AM | Report abuse

I agree 100% with natsncats. Great job, Jason, thanks!

Time to get a 1AM snack, brush and floss, and be back ready for any updates on the Euro model.

Posted by: MikeinDC2 | December 23, 2010 12:53 AM | Report abuse

Technical question:
What are the "initial conditions" that are varied for ensemble runs, and how are they varied?

Posted by: mason08 | December 23, 2010 1:00 AM | Report abuse

For the record, i am constantly searching the term "Euro model" from my house, at 1:15, because of this blog. In case anyone from a gov't agency, or my wife, thinks I am a deviant, this is my alibi. Thank God the Brazilians haven't been competitive.

Posted by: awshux | December 23, 2010 1:16 AM | Report abuse

What kind of computers are used to run these forecasts? Aren't weather models like these often what supercomputers are built for? (Or maybe climate models.)

Posted by: KBurchfiel | December 23, 2010 1:21 AM | Report abuse

Well - all of this hype this week and this is what happens? Still things can change but this bites the big one! Can we call this Dissapointmageddon?


Posted by: greg2010 | December 23, 2010 1:30 AM | Report abuse


You're right, here is what they use (and have used in the past):

Posted by: MikeinDC2 | December 23, 2010 1:43 AM | Report abuse

MikeinDC2, thanks for the very interesting page! Those old computers are a riot--you could sell them to a dance club for furniture.

Their most recent one is FAST. I wonder if anyone uses it to check Facebook during off hours ;-D

Posted by: KBurchfiel | December 23, 2010 2:12 AM | Report abuse

Well my fellow weather nerds, I hope you've now learned not to waste so much time hoping for the "big one" and analyzing every model that comes out every few hours and instead do what normal, sane, rational people do... Wait until 48 hrs before hand beforehand before you get excited.

Posted by: Thundershock | December 23, 2010 2:43 AM | Report abuse

For all those that haven't seen the latest run of all the models big storm for DC :( Sorry...once again- OTS.

Posted by: 4seams | December 23, 2010 5:29 AM | Report abuse

As my meteorology professor used to say, "after the first few days the models start getting into la-la land" and begin get progressively unrealistic (sometimes after about 6 days you'll see models predicting huge storms and gigantic swaths of precipitation that never happens).

It's rare that we're certain about any kind of specific predictions until about 48 hours before-hand -- a good rule of thumb and something that they probably should have noted on the page you linked. If you go to your local NWS forecast page you can see that they are very aware of this and limit their predictions after a few days.

Hope that helps you make sense of all this talk of models! =)

Posted by: SpeedLimit186000 | December 23, 2010 5:54 AM | Report abuse

Last night's euro model jumped east and now is clustered with the GFS and the other models in suggesting that the strong development will be offshore enough to keep us out of a major snowstorm. The last tow runs of the nws ensembles give dc a 50% chance of getting an inch or more of snow but only give a 10 to 20% chance of getting over more than 5 at least during any 24 hr period.

Posted by: wjunker | December 23, 2010 8:51 AM | Report abuse

Why would the NWS have a high wind advisory for the western suburbs but not the eastern suburbs? The wind seems pretty uniform everywhere, gusting to 35 mph at Winchester and 36 mph at Salisbury as of 9 AM.

That is trying to micromanage weather forecasting to an absurd level.

Posted by: frontieradjust | December 23, 2010 10:04 AM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2012 The Washington Post Company